INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Controversy remains regarding the effectiveness of bronchodilator agents in infants and the best techniques for measuring a physiological response to their administration. Studies using variations of the rapids thoracic compression technique have produced variable results. Some authors have found no evidence of bronchodilator responsiveness (1), while others report a "paradoxical" bronchoconstrictor response (2, 3). In contrast there are studies showing that, although inhaled bronchodilators have no detectable bronchodilator effect, they do offer protection against induced bronchoconstriction (4, 5), and speed recovery from histamine-induced bronchoconstriction (6, 7). Older studies using plethysmographic techniques do report improvements in lung function (8, 9). Some clinical studies using symptom scores have found a definite clinical benefit (10), although this finding is not universal (11). Clinically there is a consensus that bronchodilators are of value in some wheezing infants (12), although defining which infants are likely to benefit is difficult.

This apparent discrepancy between the clinical and physiological studies may be related to the techniques used to assess the physiological response. Bronchodilators may be expected to act predominantly on airways; thus, techniques that directly measure airway function may be better suited to detecting the response to bronchodilators. The flow resistance of the airways (Raw) and the mechanical status of the respiratory tissues can be estimated by using plethysmography (13), techniques using occlusions made at the airway opening (14), or measuring respiratory impedance (Zrs) by applying forced oscillations into the respiratory system (15). We recently demonstrated that an infant's lung function can be partitioned into components representing the airway and tissue mechanical properties from low-frequency Zrs data collected during Hering-Breuer reflex-induced pauses in breathing in sedated infants (15). The present study represents a pilot study to determine whether this new technique could be used to measure the effects of an inhaled bronchodilator in infants.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Twenty-two infants were recruited into the study. All were male, 13 had a history of recurrent wheeze (three or more wheezing episodes or one lasting greater than 4 wk = W), and nine had no history of respiratory illness and were classified as healthy (H). Table 1 shows the demographic details of these infants. All infants were studied at a time when they were free from respiratory symptoms. The studies were conducted with the infant supine and asleep following an oral dose of chloral hydrate (80-100 mg · kg¹). Heart rate (HR) and arterial oxygen saturation (Sa_O2) were monitored continuously (Nellcor Inc., Hayward, CA). All parents gave written informed consent for the study and were present during the tests. The study was performed with the approval of the hospital's human ethics committee.

Equipment

The technique used to collect low-frequency Zrs spectra in infants has been described in detail previously (15). This technique makes use of the brief pause in breathing that can be produced by occluding the airway with the infant's lungs inflated to a transrespiratory pressure of 20 cm H₂O, thus invoking the Hering-Breuer reflex. During this pause in breathing, the face mask was connected to a loudspeaker-in-box system and the oscillatory measurements were made. The cheeks and buccal floor were supported and mouth closed in all infants. Spontaneous respiration occurred immediately after the mask was removed from the infant's face.

The loudspeaker delivered a low-frequency pseudorandom signal containing 16 frequency components between 0.5 and 21 Hz. The relative amplitudes of the lower frequency components were elevated to maintain an optimal signal to noise ratio at all frequencies. The resulting oscillatory pressures and flows were less than 1.3 cm H₂O and 0.085 L · s¹, respectively. Oscillatory flow (V') was measured with a screen pneumotachograph (28 mm i.d.) and differential pressure transducer (ICS 33NAOO2D; ICSensors Inc., Milpitas, CA). An identical transducer sensed oscillatory pressure (Prs) as the pressure difference between the mask and a reference box, which was used to eliminate the 20 cm H₂O pressure difference between the system and the atmosphere, thus increasing the resolution of the oscillatory pressure signal. The Prs and V' signals were low-pass filtered at 25 Hz, sampled at 128/s with a 12-bit analog-digital converter, and stored on a personal computer.

Input impedance of the respiratory system (Zrs) was computed from the cross-power spectra between the measured V' and Prs signals, and the stored driving signal. The spectra were obtained from the 4-s long recordings by fast Fourier transformation from 2-s time windows with 95% overlapping. The Zrs spectra were ensemble averaged for each infant and, after omission of data points corrupted by cardiac artifacts, fitted to the following model:

(1)

where Raw and Iaw are the frequency-independent resistance and inertance attributed to the airways; G and H are the coefficients for tissue resistance and elastance, respectively; j is the imaginary unit; and  is the angular frequency raised to the power  = 2/ arctan (H/G). A shunt correction was made to subtract the impedance of the mask.

Study Protocol

Five technically acceptable measurements of lung function were made at baseline and following administration of the aerosol. Measurements with evidence of respiratory effort or leak around the mask were discarded. At each test time, the mean values of the five measurements were calculated and reported. To determine whether changes in lung function that occurred were due to the salbutamol and not due to the time after sedation (10), six infants received a placebo (Allen & Hanburys, Victoria, Australia), delivered in an identical manner. Three of these infants also received the active drug.

Five actuations from a salbutamol metered-dose inhaler (100 µg per actuation, Ventolin, Allen & Hanburys, Victoria, Australia) or placebo aerosol, which was shaken between each dose, were delivered through a Nebuchamber (Astra, Sweden) with a tightly fitting face mask. The Nebuchamber is a small-volume metal holding chamber with a separate inlet and outlet valves devised for young children (16). The relatively high dose of salbutamol was selected in order to ensure adequate delivery of drug to the lungs to produce a response in line with other similar studies (4). Five breaths were allowed to ensure that the chamber had been fully emptied between actuation. Five minutes was allowed for the drug to take effect prior to retesting.

Statistical Analysis

Two-tailed paired t tests were used to detect a significant effect of salbutamol on the measures of lung function. The significance of this change compared to the placebo group was determined by performing an unpaired t test on the residuals (before value minus after value) of the active and placebo groups. The same method was used to look for significant differences between the healthy and wheezy groups. In order to avoid bias due to multiple tests of significance, (0.05) was adjusted according to the formula: 1 = 1  ([1  ]^[1/n]) where 1 is the adjusted level of significance and n is the number of t tests performed in the analysis. Thus, when six t tests are required, the appropriate level of significance is 0.008.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Measurements of Zrs were possible in all 22 infants and were of sufficient quality to allow Raw, Iaw, G, and H to be calculated, with an average fitting error of the model to the data of less than 3%. The baseline mechanical parameters did not differ between the normal infants and those with a history of recurrent wheeze (Tables 234).

Heart Rate and Sa_O2

Following salbutamol inhalation, a 16% rise was seen in heart rate (p < 0.008). There was no increase in heart rate following placebo. A small but statistically significant fall was seen in Sa_O2 both after placebo (1.3%) and after salbutamol (1.1%).

Lung Function

Following inhalation of salbutamol, all infants showed a reduction in Raw, which fell by 11% for the group as a whole (p < 0.008) (Table 2, Figure 1). While the fall in the infants with recurrent wheeze (17%) was greater than that in the normal group (8%), this difference did not reach statistical significance (p = 0.12). There was an 8% fall in G following salbutamol (p = 0.05) for the group as a whole, although there was more variability in individual responses with 4 infants showing slight increases (Table 3). There was no difference between the magnitude of the fall in G between the normal infants (13%) and those with recurrent wheeze (6%) (p = 0.57). There was no systematic pattern for a change in H following salbutamol inhalation, either for the group as a whole or for either subgroup (Table 4). Although only a small number of infants were studied, there was no suggestion of any systematic change in any lung function parameter following the placebo aerosol (Tables 234).

View larger version (29K): [in this window] [in a new window]   Figure 1.   Plots of individual values for airway resistance (Raw), airway inertance (Iaw), tissue resistance (G), and elastance (H) before (pre) and after (post) salbutamol (open symbols) or placebo (filled symbols). Wheezy (squares) and healthy (triangles) infants. Significance tested with an unpaired t test of active treatment against placebo.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In a recent study (15), we demonstrated that the pause in breathing produced by invoking the Hering-Breuer reflex gives suitable conditions for collecting reliable, low-frequency Zrs data in sleeping infants. We also showed that, using an appropriate model to characterize Zrs, lung function can be partitioned into components representing the mechanical properties of the airways and lung tissues. In the present study we have demonstrated that this technique is sensitive enough to assess the effects of salbutamol on airway and tissue mechanics in infants. Following salbutamol administration, we found a moderate but statistically significant decrease in Raw, with no significant changes in lung tissue parameters, indicating that salbutamol acts predominantly to increase airway caliber in infants.

There is ample evidence from studies in adults and older children (17, 18) that the FOT, applied to a higher frequency range than used in the present study, detects bronchodilator-induced changes in respiratory resistance (Rrs), although its sensitivity in subjects with severe airway obstruction has been challenged (19). As these studies used frequencies above 4-6 Hz, their measurements of Rrs will predominantly reflect Raw, with little to no contribution from the lung tissue mechanical properties. No studies prior to the present one have shown a similar response in infants with this technique, although changes in Rrs have been detected with the passive flow volume technique (20) and with plethysmography (21). The changes we detected in Raw were clear and were statistically significant. While the number of infants receiving placebo was small, none showed a fall in Raw following placebo. This observation suggests that the falls in Raw seen following salbutamol were due to the bronchodilator action of the drug and not due to changes in the level of sedation or an effect of order (an effect due to the order that the measurements were made rather than the intervention).

All of the above techniques, including ours, include the resistive properties of the glottis and upper airways in their measurement of resistance. Thus, we are unable to exclude changes in upper airway resistance as a contributor to the decrease in Raw. However, as our measurements are made during a Hering-Breuer reflex-induced pause in breathing at a trans-respiratory pressure of 20 cm H₂O, the upper airway is likely to be held open and unlikely to contribute to the fall in Raw. In addition, the absence of breathing movements during our measurements exclude any possible influence of salbutamol on breathing pattern (22) as an explanation of our results. We can conclude, however, that the fall in Raw is due to the actions of salbutamol in the respiratory tract, presumably on airway smooth muscle.

The importance of tissue resistance (Rti) in lung mechanics is well recognized from animal studies (23). There is also evidence that salbutamol decreases Rti in puppies (24). Measurements of peripheral airway resistance include lung tissue resistance and have been made in adults (25). These workers have developed a technique for partitioning lung resistance into components representing the central airways (Rc) and the peripheral airways (Rp) by wedging a catheter-tipped micromanometer (CTM) into 3-mm airways of human volunteers. Their measurements of Rp include the resistance of the pulmonary parenchyma. Yanai and colleagues (26) have demonstrated that the "peripheral airways" are the predominant site of airflow obstruction in adults with asthma, chronic bronchitis, and emphysema. Furthermore, using the CTM, Ohuri and coworkers (25) demonstrated that inhaled atropine sulfate dilated the central, but not peripheral, airways, whereas inhaled beta-adrenergic agents dilated both central and peripheral airways. In the present study we found evidence suggestive of an effect of salbutamol on G, our measure of Rti, but although this effect had an associated p value of < 0.05, it did not reach formal significance at the corrected 0.008 level of 1. However, studies in animals breathing a neon-oxygen gas mixture (27, 28) have suggested that changes in Raw and in G following a challenge with a constrictor agonist, which are not accompanied by parallel changes in H, are likely to be due to the development of heterogeneous constriction inducing ventilation inhomogeneity. One argument against that being the explanation for the results of the present study is that the changes in G were also seen (and if anything, to a greater degree) in the normal infants. Thus, either significant ventilation inhomogeneity exists in normal sedated infants and is reversed by salbutamol, or inhaled salbutamol has actions on the tissue dissipative (or resistive) mechanical properties. In either case, this preliminary observation warrants further study.

In the present study, we found no evidence of an effect of salbutamol on Iaw or H. Airway inertance was affected by salbutamol in individuals but not in any consistent manner. The importance of Iaw in the frequency range studied here is doubtful, and its contribution is ignored in most lung models. Dynamic respiratory compliance (Cdyn) has previously been shown to be affected by salbutamol (17). The lack of effect on H could be explained by differences in methodology in our study from previous techniques. First, Cdyn, as used in previous studies, includes a component of resistive pressure losses in the lung tissues as well as elastic pressure losses. Second, as our measurements were made during a Hering-Breuer reflex-induced apnea, any changes in breathing pattern resulting from salbutamol administration (22) will not affect our measurements but can have marked influences on measurements of Cdyn.

The increase in HR following salbutamol, but not placebo, suggests that a significant dose of salbutamol was delivered from the metal spacer to the infant's lungs. It is interesting to note that, when compared to baseline, the fall in Sa_O2 that occurred in response to salbutamol was statistically (but not clinically) significant; however, a similar fall also occurred with the placebo inhaler. This is consistent with a previous report suggesting that it was caused by the chloral hydrate used to sedate the infants (29), although in that study the falls were greater and the infants were hospitalized with acute viral bronchiolitis and were studied before discharge.

A major methodologic difference between our study and previous studies in infants is the fact that during our measurements, the lungs are held inflated at a transrespiratory pressure of 20 cm H₂O. This would have the effect of splinting open the airways during the measurements, especially as the transmural pressure is being applied from "inside" the airways. Airway caliber is a complex function of lung volume, transmural pressure, and airway wall compliance. Salbutamol functions to increase airway wall compliance by relaxing airway smooth muscle. Thus, under our measurement conditions, an increase in airway wall compliance would be expected to result in an increase in airway caliber, as lung volume and transmural pressure are constant, and a decrease in Raw. This is in stark contrast to what would be expected to happen during a forced expiration produced by an external pressure, as is used in the rapid thoracic compression technique. Here, the same increase in airway wall compliance may be expected to result in a smaller airway caliber during the forced expiration as the airway is "squeezed" from outside. This may cancel out any increase in resting airway caliber resulting from relaxation of airway smooth muscle overcoming bronchoconstriction. This same phenomenon has been reported in older children with cystic fibrosis (30).

The technique used in the present study is simple to implement and does not require complicated equipment. We used a low-pressure pump to inflate the lungs to a transrespiratory pressure of 20 cm H₂O and a balloon valve to occlude the airway, using a system that was designed for measuring forced expiratory parameters from raised lung volumes (31). However, this equipment is not necessary, and the lungs can easily be inflated using a resuscitation bag, with the airway being occluded by a hand-operated slide valve (32). The oscillatory signal is produced by a computer-driven loudspeaker (15). We have found that both normal infants and those with a history of recurrent wheeze reliably produce Hering-Breuer reflex- induced pauses in breathing efforts long enough (6 s) to allow measurements of Zrs. Our studies are routinely performed under sedation with chloral hydrate, given in a dose of 80-100 mg/kg for infants more than 1 mo old, under the direct supervision of a pediatrician. The studies all received approval from our institutions' human ethics committees, and parents gave written informed consent. Parents are encouraged to stay and witness the study and most do so. There is no doubt that the sedation is the component that parents are most apprehensive about before the study. However, following the study, parents report that the study has not distressed their children and is not associated with problems at home after leaving the laboratory (data not shown). More than 2,000 studies in normal and wheezy infants have been performed in our laboratory during the past 15 yr under chloral hydrate sedation without incident.

The clinical relevance of our findings are uncertain. The primary purpose of the present study was to determine whether the measurement of Zrs using the low-frequency FOT was sensitive enough to detect a bronchodilator response in infants. Based on the variability of the measurements reported in Table 2, four infants would be required to detect a 20% change in Raw with 95% power at the 5% level, using a paired t test. Thus, the test is extremely sensitive. This is also demonstrated by the fact that statistically significant decreases in Raw were shown in both healthy and wheezy infants. However, the power of the test to detect the difference between two groups of infants is considerably less. Based on the responses of the healthy and wheezy infants to inhaled salbutamol, the study had only approximately 25% power to detect a difference between the groups. To have 80% power to detect a difference in the response between the groups, approximately 25 infants per groups would be needed. However, despite their past medical history, none of the wheezy infants were symptomatic at the time of testing, and it would be unusual to find bronchodilator responses in mild childhood asthmatics when they were asymptomatic. Nevertheless, as a population, normal adults will show increases in FEV₁ and decreases in Raw following inhaled bronchodilator. A change in FEV₁ of up to 190 ml is within the 95% confidence limits for repeated tests in adults (33).

In summary, the results of the present pilot study demonstrate that the low-frequency FOT is a suitable method for studying the vexing question of whether infants do respond to inhaled bronchodilators. Furthermore, this technique has the potential to contribute substantially to our knowledge of respiratory physiology in infants.